A Multi‐Method Geospatial Framework for Drought Risk Assessment in the GAP Region (Türkiye): Trend Analysis, Dry‐Spell Persistence and Random Forest Forecasting
- 1Department of Geography Nakhchivan State University Nakhchivan Azerbaijan
- 2Geography Department, Faculty of Arts and Sciences Igdir University Igdir Türkiye
Abstract
This study presents a multi‐method analysis framework that combines trend analysis, dry period persistence and machine learning‐based projections to comprehensively assess drought risk in the Southeastern Anatolia Project (GAP) Region. The study first identified trends for the 2000–2024 period using the Palmer Drought Severity Index (PDSI) using the Mann–Kendall and Theil–Sen methods. Then, the run theory approach was used to determine the length, frequency and persistence of dry periods. Finally, hydroclimatic projections were generated for the 2025–2030 period using the random forest regression model. The findings indicate that long‐term PDSI trends in the region are generally weak, but dry period durations are quite long (average 15–22 months), with maximum values reaching up to 70 months. This suggests that drought in the GAP Region is structural and persistent. The random forest model achieved high explanatory power ( R 2 = 0.8666) on the test data and successfully reproduced the region's strong seasonality. Future projections indicate a significant recurrence of summer drought across all provinces, with some provinces, such as Gaziantep and Kilis, prone to drier conditions. Overall, the study presents a holistic methodological approach that considers past trends and drought persistence alongside future climate signals, providing strong scientific evidence for regional water management, agricultural planning and climate adaptation strategies.
Related
- Hybrid machine learning framework for drought forecasting: integration of LSBoost, RQNN, CNN-RNN, CNN-LSTM and advanced data decomposition
- Assessment and forecasting of drought in Nepal using long short-term memory model
- Drought and flood risk assessment under the IPCC risk framework in an agriculturally dominated tropical river basin, India
- An intelligent weighted adaptive regression process for reliable and uncertainty aware drought forecasting
- An Artificial Intelligence Integrated Geospatial Decision-Support Framework for Scenario-Based Climate-Adaptive Land-Use Optimization and Environmental Risk Reduction
BibTeX
@article{celik2026multimethod,
title = {{A Multi‐Method Geospatial Framework for Drought Risk Assessment in the GAP Region (Türkiye): Trend Analysis, Dry‐Spell Persistence and Random Forest Forecasting}},
author = {Mehmet Ali Çelik},
journal = {Ecohydrology},
year = {2026},
month = jun,
doi = {10.1002/eco.70234},
url = {https://doi.org/10.1002/eco.70234},
note = {CC BY},
}
RIS
TY - JOUR
TI - A Multi‐Method Geospatial Framework for Drought Risk Assessment in the GAP Region (Türkiye): Trend Analysis, Dry‐Spell Persistence and Random Forest Forecasting
AU - Mehmet Ali Çelik
PY - 2026
DA - 2026/06/01
JO - Ecohydrology
DO - 10.1002/eco.70234
UR - https://doi.org/10.1002/eco.70234
AB - This study presents a multi‐method analysis framework that combines trend analysis, dry period persistence and machine learning‐based projections to comprehensively assess drought risk in the Southeastern Anatolia Project (GAP) Region. The study first identified trends for the 2000–2024 period using the Palmer Drought Severity Index (PDSI) using the Mann–Kendall and Theil–Sen methods. Then, the run theory approach was used to determine the length, frequency and persistence of dry periods. Finally, hydroclimatic projections were generated for the 2025–2030 period using the random forest regression model. The findings indicate that long‐term PDSI trends in the region are generally weak, but dry period durations are quite long (average 15–22 months), with maximum values reaching up to 70 months. This suggests that drought in the GAP Region is structural and persistent. The random forest model achieved high explanatory power ( R 2 = 0.8666) on the test data and successfully reproduced the region's strong seasonality. Future projections indicate a significant recurrence of summer drought across all provinces, with some provinces, such as Gaziantep and Kilis, prone to drier conditions. Overall, the study presents a holistic methodological approach that considers past trends and drought persistence alongside future climate signals, providing strong scientific evidence for regional water management, agricultural planning and climate adaptation strategies.
SN - 1936-0584
SN - 1936-0592
ER -